A positive temperature coefficient sensor (PTC) in embedded power design is far more than a simple resettable fuse. Whether you are using a polymer PTC (PPTC) for overcurrent protection or a ceramic switching PTC for direct thermal cutoff on a motor driver, the component's behavior is entirely dictated by the thermal path from the silicon junction to the ambient air. If you misjudge the thermal resistance, your PTC will either nuisance-trip under normal load or fail to protect the circuit during a fault. Here is how to mathematically size your thermal management system and integrate a PTC sensor effectively.
Thermal Path Math: Junction to Ambient (Rθ) with PTC Trip Points
To keep your power components alive, you must calculate the total thermal resistance from the semiconductor junction to the ambient environment ($R_{\theta JA}$). The fundamental equation governing this heat flow is:
$T_J = T_A + P_D \times (R_{\theta JC} + R_{\theta CS} + R_{\theta SA})$
Where:
$T_J$ = Junction temperature (°C)
$T_A$ = Ambient temperature inside the enclosure (°C)
$P_D$ = Power dissipation (Watts)
$R_{\theta JC}$ = Thermal resistance, junction-to-case
$R_{\theta CS}$ = Thermal resistance, case-to-sink (interface material)
$R_{\theta SA}$ = Thermal resistance, sink-to-ambient (the heatsink)
Worked Example: Sizing for an IRLB3034 MOSFET
Let’s say you are switching a 12V, 10A heater load using an IRLB3034 (TO-220 package). The MOSFET has an $R_{DS(on)}$ of roughly 3mΩ at $V_{GS} = 10V$. Power dissipation ($P_D = I^2 \times R$) is $10^2 \times 0.003 = 3W$.
Your enclosure ambient ($T_A$) is 40°C. You want to keep the junction ($T_J$) below 125°C to maintain a safe margin. The datasheet specifies $R_{\theta JC} = 1.5°C/W$. Using a standard silicone thermal pad, $R_{\theta CS} \approx 0.5°C/W$.
Plugging in the numbers:
$125 = 40 + 3 \times (1.5 + 0.5 + R_{\theta SA})$
$85 = 3 \times (2.0 + R_{\theta SA})$
$28.33 = 2.0 + R_{\theta SA}$
$R_{\theta SA} = 26.33°C/W$
You need a heatsink with a thermal resistance of 26.33°C/W or lower. A bare TO-220 in free air has an $R_{\theta JA}$ of about 62°C/W, which would push the junction to 226°C and destroy the silicon. For a deeper dive into thermal resistance networks, All About Circuits provides an excellent breakdown of the electrical analog model.
Heatsink Selection and PTC Placement Strategy
Knowing you need an $R_{\theta SA}$ of 26.33°C/W allows you to select a real, off-the-shelf heatsink. Here is how two common Aavid Thermalloy (now Boyd Corporation) extruded aluminum heatsinks compare for this exact TO-220 application:
| Part Number | Type | $R_{\theta SA}$ (Natural) | Verdict for 3W Load |
|---|---|---|---|
| 577202B00000G | Low-profile extruded | ~24.0°C/W | Pass. Keeps $T_J$ at ~112°C. |
| 530002B02500L | High-fin staggered | ~5.0°C/W | Overkill. Keeps $T_J$ at ~55°C, but wastes board space. |
| No Heatsink | Bare TO-220 | ~62.0°C/W | Fail. $T_J$ exceeds 200°C. |
Where to Mount the Positive Temperature Coefficient Sensor
If you are using a ceramic switching PTC (like the TDK EPCOS B59100 series) to act as a hard thermal cutoff for the gate drive, physical placement is critical. Do not mount the PTC on the PCB near the MOSFET. Instead, use a thermally conductive epoxy (like Arctic Alumina) to bond the disc PTC directly to the exposed metal tab of the TO-220. The tab is electrically tied to the drain, so if your circuit topology requires isolation, use a Kapton tape layer between the PTC and the tab, accepting a slight penalty to $R_{\theta CS}$.
Interpreting Derating Curves and Airflow Dynamics
A common mistake in embedded design is assuming a PPTC (Polymer Positive Temperature Coefficient) resettable fuse will trip at its rated current regardless of the environment. PPTC hold currents are strictly derated based on ambient temperature. A Littelfuse 1206L050YR rated for a 0.50A hold current at 20°C might only hold 0.30A at 60°C. If your enclosure lacks ventilation, the ambient heat alone will push the PPTC closer to its trip threshold, causing nuisance tripping during normal operation.
What airflow buys you: Moving from natural convection to forced air (even a small 30mm 5V brushless fan pushing 4 CFM) can cut a heatsink's $R_{\theta SA}$ by 50% to 70%. If our 577202B00000G heatsink drops from 24°C/W to 10°C/W under forced air, the MOSFET junction temperature drops from 112°C to 70°C. This massive thermal headroom allows you to use a smaller, cheaper PTC sensor with a lower trip threshold, or safely increase your continuous load current.
FAQ: Positive Temperature Coefficient Sensor Integration
How hot is too hot for a PTC sensor before it degrades?
It depends on the PTC chemistry. Polymer PPTCs (the resettable fuses) typically have a maximum continuous operating temperature of 85°C. Exposing them to sustained temperatures above this, or repeatedly tripping them in high-current fault conditions, causes the polymer matrix to degrade, leading to a permanent upward shift in their base resistance. Ceramic switching PTCs (Barium Titanate based) can handle much higher temperatures, often up to 150°C or 175°C, but exceeding their specific $T_{max}$ rating will cause the ceramic grain structure to crack or the silver electrodes to delaminate.
What airflow or enclosure changes buy you more thermal headroom?
Beyond adding a fan, you can increase the enclosure's effective surface area by using the chassis itself as a heatsink (mounting TO-220s directly to an aluminum backplate). If you must use a sealed plastic enclosure (IP65 rated), you cannot rely on airflow. In sealed environments, you must derate your power components by at least 30% and select a PTC sensor with a trip point 15°C lower than you would in an open-air design to account for the trapped stagnant air boundary layer.
What are the physical failure signatures of thermal stress on a PTC?
When a PTC is subjected to chronic thermal stress or excessive fault energy, look for three signatures:
1. Epoxies and Coatings: The outer epoxy coating will yellow, micro-crack, or bulge.
2. Solder Joints: The leads will show signs of solder fatigue (a dull, crystalline ring around the lead exit) due to repeated thermal expansion and contraction during trip/reset cycles.
3. Electrical Shift: The post-trip resistance ($R_1$) measured one hour after a trip event will be significantly higher than the datasheet specification, indicating internal matrix damage.
How does a switching PTC differ from a silicon PTC for microcontroller ADCs?
A ceramic switching PTC has a highly non-linear resistance curve; it sits at a low resistance (e.g., 100Ω) until it hits its Curie temperature, then spikes to 10kΩ+ almost instantly. It is useless for proportional temperature measurement. If you want to read exact temperatures via an ESP32 or Arduino ADC, you need a Silicon PTC sensor (like the Infineon KTY81-220). Silicon PTCs have a nearly linear positive temperature coefficient over a wide range (-50°C to +150°C), making them ideal for voltage-divider circuits feeding a microcontroller's 12-bit ADC for closed-loop fan control.
